31 replies to this topic

[Grippen]

Hello! I got Two devices, Octagon SF8008

 

Same problem on both of them..  I Will try to explain.
When I wants to reboot the box from menu i press button  "Reboot" but the box doesn't reboot it freezes and tv is saying "NO SIGNAL" - BUT some times , it works, but next time I try it does not work..

it's the same from SSH/Telnet.

 

I have try this OpenPli images

openpli-8.3-release-sf8008-20230129_usb
openpli-8.3-release-sf8008-20230219_usb

openpli-8.3-release-sf8008-20230226_usb

openpli-8.3-release-sf8008-20230305_usb

I have try some other images - (Not Openpli) and there the reboot works like it should. 

 

but i really want to use Openpli.

 

Is it some fix? or is it just me that's have this problem?

 

 

Best Regards

Edited by Grippen, 13 March 2023 - 12:36.

[neo]

NFS mount active?

[Grippen]

NFS mount active?

 

Hello Thanks for answer! , No I don't use NFS mounts on the box, 

When i try different pli images everything was "original" just installed the images and try to reboot box 3-4 times

 

Best regards 

 

[Pr2]

Hello,

 

Do you have any USB device(s) attached to the box?

Do you have an internal HDD?

Is there an MMC card plugged in?

[Grippen]

Hello,

 

Do you have any USB device(s) attached to the box?

Do you have an internal HDD?

Is there an MMC card plugged in?

Hello Thanks for answer!  - No Usb , No Hdd , No MMC card - Nothing attached to the USB ports and nothing internal either

 

Best Regards 

[neo]

If you leave it alone, does it restart after a while, say 5 to 10 minutes?

[Grippen]

If you leave it alone, does it restart after a while, say 5 to 10 minutes?

I Can try this right now

[neo]

Others have reported the same issue, but say that eventually it will boot through.

 

We've always throught it was related to mounts over wireless (are you using wifi or a wired network connection?), but if you don't use any mounts?

Edited by neo, 15 March 2023 - 15:19.

[s3n0]

Hi.

 

First, connect via FTP connection to your set-top box and check the contents of the crash-log files (you can find them here: /home/root/*crash*.log).

 

You can follow the date and time, which should be hidden in the name of these LOG files. If there is no date and time in the file name (I don't remember... sorry), then follow the numbers from the file name - the larger the number, the newer the file. I think these numbers represent the UNIX Time Stamp (counting from 1/1/1970, 00:00:00) and if you'd like, here's a converter: https://www.unixtimestamp.com/

 

If you found any current crash-log files by date and time in the file name, you can attach them here as a file attachment.

 

------------------

 

If you did not find any crash-log files in the set-top box (with the current date and time in the file name), you can also try debugging Enigma2. However, when the problem is not in Enigma2 but in the operating system, this debugging probably won't catch the error anyway.

 

Restart your OpenPLi in debug mode - every time you need to test your problem:
https://wiki.openpli...g_from_the_menu

 

Or better via Linux Shell, which is a faster solution:
https://wiki.openpli...he_command_line

(for example, use a terminal connection to the set-top box, via SSH or Telnet protocols, with the help of the PuTTY application)

 

After a new start of OpenPLi, the debug-log file will be automatically generated. Then trigger an error - i.e. attempt to restart the set-top box. If in this case the error really manifests itself, then when OpenPLi is started again, find the created debug-log file and also the crash-log file (if you find one there) - according to the date and time of the approximate start of the set-top box. And put them here - in the discussion. It is ideal to copy these log files via an FTP connection. You can find them in the set-top box in the directory: /home/root

[neo]

Chances are there are none.

 

This issue is known (reboot delay in combination with a wireless network connection and an active mount), but the cause isn't.

 

As it does reboot after a while, it must be something that waits on a timeout, we suspect the network connection is dropped before the shares are umounted.

 

But weird that is only happens on the SF8008...

Edited by neo, 15 March 2023 - 17:51.

[s3n0]

Problems with NOT rebooting could be anywhere else.

 

Of course, this ONE AND ONLY possibility, which you mention and which has not yet been substantiated by anything (further clues) from the interviewer, is not ruled out either. Or did you invent and add clues yourself that would determine the cause of this particular problem ? The problem of NOT rebooting can also be due to one of hundreds of other possible causes. Just because you believe it based on one fact doesn't mean it's the real cause of the problem. If at least 2-3 other clues lead to the problem, I recognize that it can be caused with a high probability by what you mentioned above. But since these clues are missing, we do NOT know where the problem is and we MUST find it.

 

If it turns out that the problem was in what you mentioned above, then it will only be a 1 to 1 coincidence.

 

Where exactly did the interviewer write something about network malfunctions or the use of WiFi ? Where exactly did he write something about using an external network drive ? I must have missed it somehow. Can you copy it for me please ? Thanks.

 

[neo]

I didn't write what, I asked the TS about that in post #8.

 

The TS does mention that the TV reports NO SIGNAL, which means the box has already shut off the HDMI.  The TS also reports that the problem also occurs when rebooting from the commandline.

 

Which are all indications of the problem already observed by others.

 

I agree that we must find it, but as the problem occurs during linux shutdown, you will not find it with enigma debugging, as it has nothing to do with Enigma, the same happens with "init 4; sleep 30; reboot".

Edited by neo, 15 March 2023 - 21:39.

[rantanplan]

https://github.com/o...n_4.4.35.bb#L53

 

https://github.com/o...00_reboot.patch

[neo]

Interesting...

[el bandido]

@Grippen,
You may try the OpenPLi Homebuild image in the link below. This image has the changes or additions shown in post #13 of this thread. Maybe your Reboot problem is fixed.
 

http://162.216.113.2...ot_ofgwrite.zip

Edited by el bandido, 16 March 2023 - 06:27.

[Grippen]

Hello! oh Many answers, I will try to answer everyone of you guys! 

 

Sorry for late answer from me! 

 

neo

 I Try to reboot the box yesterday, and it worked for two times, but on the third time, the box freeze's, but I left it in "freeze mode" for almost 10 hours, and the box did not reboot, I had to kill it with the ON / OFF switch..

I use wired network connection, No Wifi connected. And no Mounts.

 

 

 

s3n0

I cannot find any Crash logs, Should I try with enigma debug mode? - but It maybe dosent show anything?

 

neo X2

When I try command - 

init 4; sleep 30; reboot - it will also freeze 

 

rantanplan

is this some fix?
 

el bandido

I get back to you when I have tested

 

 

Best Regards! 

Edited by Grippen, 16 March 2023 - 22:53.

[Grippen]

el bandido

I Have try this home made image right now , an it also Freezes when I press REBOOT :/

 

 

Best Regards!

[rantanplan]

mh..
As I understand it, you have built a home build image with the patch.

I just pointed out the commit regarding your reboot problem.
According to this, 2 manufacturers have the problem, because the gb-trio4k is also affected.
According to the commit it is due to the internal flash being used.
Therefore, not all receivers have it, only certain series.

The question that remained was whether the image of 'el Bandido' hadn't applied the patch yet?

Probably your versions of the other images were also current (python3) versions?

It's a bit difficult to port the patch back to 8.3.

@neo
in Octagon develop should I update the bsp layer with the patch?
As a test?

[neo]

Both the GB Trio 4K and the uClan Ustym4Kpro are hardware identical to the SF8008.

 

It would probably be better if someone that actually has one can test the patch before including it. It's better than commit and revert imho .

[rantanplan]

A simpler way would be to insert patch outside BSP in develop as bbappend.

The user can then test it.
8.3 could also be used in this way backport.

If the test is successful.